Sharing data between research centers is increasingly important for contemporary brain imaging studies because they involve large numbers of subjects and complex analysis protocols that require highly specialized expertise. Our long-term objective is to facilitate brain-imaging research by enabling remote researchers to pool data between institutions and to analyze data using the appropriate algorithms executing on distributed resources. There are a number of difficult data management and technology challenges that have limited the success of data sharing environments. Rather than attempt to develop a comprehensive and general solution, we propose to develop a set of open, interoperable, and portable software tools that address critical issues currently limiting efforts to share and analyze brain-imaging data. Building upon years of providing brain-mapping expertise to collaborators, we propose to solve problems that we repeatedly encounter and that currently limit progress in brain imaging research. We propose to develop validated tools that enable collaborators to remotely access a variety of data analysis methods and databases. We will create web-based tools to perform multi-institutional studies, and provide access to complex data processing protocols executing on distributed computing resources. There are three specific aims. 1) Enable the web based acquisition and management of data utilizing an access control system that includes consideration of subject consent limits and investigator imposed conditions to facilitate data pooling for multi-institutional studies. This system will convert data files between different formats and schemas so that data can be used consistently between analysis programs and databases. It will also anonymize images and metadata according to institution-specific protocols. 2) Develop a system that is aware of data type and provenance so that it may act intelligently to arbitrate between different analysis programs. This system will capture the expertise of experienced lab personnel in the usage of various tools and assist new users in designing appropriate analytic strategies. 3) Create meta-algorithms that improve the robustness of techniques for neuroimaging analysis by intelligently combining the results from multiple algorithms. The proposed approach will provide a set of tools that address significant problems in data sharing and utilization. The resulting information technology will be scalable and applicable to other scientific data sharing problems. [unreadable] [unreadable]